their opposing arguments by those who favor or oppose using observational studies alone for drawing conclusions about causality. Hill's considerations are often listed as criteria to fulfill despite the fact that he specifically cautioned against doing so. His argument was that the more types of evidence available the more likely one was to being able to conclude causality from a relationship. Hill also wrote that the degree of evidence needed to get people to give up sugar and fat should be stronger than replacing a chemical in industrial use when a safer was available. Hill's position has fostered a deal debate and thinking about these issues for decades.
Among observational studies, case-control studies are now also considered flawed for assessing causality because of the differential recall of diet between cases recently diagnosed with a serious disease and that of controls who do not have that condition. That leaves primarily prospective cohort studies and there is a belief that such studies can predict cause and effect by those in nutritional epidemiology although that view has been challenged by many outside that discipline. Cohort studies are generally designed so that participants usually have no diagnosed serious illnesses and have diet assessed at the beginning of a study, and sometimes at regular intervals during extended follow-up. This latter distinction is important because those studies that apply only a single determination of dietary intake assume that the diet of participants does not change over the duration of the follow-up which is frequently 10-20 years when chronic diseases are evaluated: shorter term follow-up studies abound in the literature but are of little value because of the possibility of reverse causality that is, the presence of